Reigniting the ETL vs. ELT Debate - Which Better Supports the Business?
summary
Argues that cloud data warehouses change the ETL vs ELT choice: traditional ETL routes all bits through a beefy transform server while ELT loads fast into Redshift or Snowflake and pushes transforms into scalable compute, trading ETL server cost for cloud compute spend. Cites tools and roles: Informatica/Talend (ETL), Fivetran/Stitch (loading), dbt/Matillion (in-warehouse transforms), HighTouch (reverse ETL), plus legacy Teradata/Hadoop realities. Recommends a platform that covers loading, transforming, reverse ETL, APIs and cataloging.
AI-written summary · about
chapters (8)
- 0:00Introduction and history
- 7:36Cloud data warehouse evolution
- 13:27Modern data architecture challenges
- 18:26ETL vs ELT deep dive
- 24:05Pros and cons of ETL and ELT
- 29:02Choosing the right tools
- 35:25SnapLogic platform overview
- 37:12Closing remarks
related talks
16mETL vs ELT: Why the T has shifted
01h 28mA Guide to the Modern Data Stack
29mETL and ELT, What's The Difference?
15mNext Gen ETL Generators
41mModern Data Architectures – What They Are and How to Implement?
39mLeaving ETLs Behind: Moving from a ‘Code Less’ Solution to a ‘Code Full’ Approach
38mThe Rise & Downfall of the Data Engineer REVISITED
55mThe ELT pattern in action with DBT and SQLMesh